Magazine

Tracking images on the web

The first web pages were text-only documents. The only beautification was through formatting and colouring. Fortunately, images were not far behind and were, in turn, followed later by audio and video.

However, even though we think of the web as being a multimedia experience, the truth is that the majority of content uploaded every day is still text-based. This slant shows up in everything we do on the web, especially how we search it.

Because, even when we’re performing an “image” or a “video” search on our favourite site, we’re doing it using text. Though text-based searching is remarkably efficient at directing us the content we want, it makes it very difficult to track usage of visual works as they are copied and circulated across the web.

That, in turn, hurts the ability of content producers to fight misuse of photographs or enforce licensing agreements. This makes finding new techniques to track down image reuse critical for journalists or organizations which sell or distribute photographs.

Fortunately, there are tools and techniques that can meet that challenge.

Free Techniques

Popular image search engines such as Google Image Search  do a remarkable job finding images based upon a description. They work not by searching the images themselves, but the text around them.

However, if an image is pasted into another site, the owner has very little control over the text that is put around it. So, to make searching for the image easier, we have to attach text to it that will likely be copied with it.

The most obvious opportunity to attach text is the file name itself. Most image search engines can easily search for file names and, if your image has a unique enough moniker, it can be a great way to detect reuse as few people bother to change the name when putting it on their own site.

The easiest way to use this technique is to create system for naming your images that involves a using a unique string of letters and numbers, also known as a fingerprint.

For example, a photographer named “John Smith” might use the digital fingerprint j15smith73, which, as of right now, turns up no documents in

Google, He would then append that to his file name, creating a name such as photographinfo-j15smith73.jpg where the first part is a unique identifier for that photograph.

Then, to find pages with copies of his photographs, all the photographer has to do is search for “j15smith73” in any image search and follow through on any results that are not from approved sites.

The caveat to this system is that file names can easily be changed. That is why, as a backup, the string can also be embedded into the

Exif data contained within the image. This can be easily achieved in almost any image editor.

Though not all image search engines access Exif data, most of the big ones do and Exif data is much less likely to be modified than a file name since few know to even look for it.

Another interesting use is to combine the use of the fingerprint with

Google Alerts to automatically update you every day of new appearances of your photographs. This prevents you from having to do searches manually, saving you time and helping you do a more thorough job by breaking it into manageable parts.

However, the best feature of this system is that, unlike watermarking or other image modifications, fingerprinting allows photographers to monitor their images without obscuring them. In fact, ordinary web users will be unaware that control measures are in place at all.

Paid Alternative

Though the digital fingerprint system is useful for monitoring large amounts of photographs for free, it is far from perfect. Photographs can be edited, file names can be changed and Exif data can be wiped. Furthermore, relying on traditional search engines can be very slow and the lack of structure can make it very cumbersome to manage many cases at once.

Those with either very high value photographs or a large quantity of works to manage may require a more robust solution. A company known as

Digimarc has a service that may work.

Digimarc’s ImageBridge service adds a watermark to images that, while invisible to the naked eye, is detectable by their search spiders, which crawl the web constantly looking for and reporting unauthorized usage of photographs.

In addition to providing tracking, ImageBridge also provides proof of ownership since each watermark is also embedded with information about the owner of the photograph as well as any relevant transaction information.

ImageBridge is a popular service among stock photographers as well as wire services, freelance journalists and others that sell and distribute images for a living. For those with fewer needs, they also offer a cheaper MyPictureMarc service  that has many of the same functions but at a reduced cost.

Conclusions

Unlike text, which can be searched for with nothing but a quick trip to Google, detecting image copying requires both pre-planning and forward thinking.

However, even though the Internet was never designed to make searching for images easy, it can work if one takes a few moments to make their images easy to locate.

Unfortunately, far too many media organizations do not take the time to follow such reasonable precautions and, even as they valiantly defend the work of their writers, they let abuse of the images go unchallenged. This not only allows a great deal of abuse to thrive, but also costs them many potential licensing opportunities.

With all that is available to be gained and lost on the web, it makes sense to take any reasonable precautions available, especially if it is as simple as changing a file name.

Tags:

Related articles

• Home |
• Seminars  |
• Events |
• Media Landscape |
• Newsroom|
• Media News|
• Resources

Terms of service and Privacy policy.

Copyright © 1992- 2011 European Journalism Centre. All rights reserved.

Valid XHTML and CSS.

European Journalism Centre

Sonneville-lunet 10
6221KT Maastricht
The Netherlands

Tel. : +31 43 325 40 30
Fax : +32 2 400 42 20

E-mail: info[at]ejc.net